Armour Values

[SFraser]

You have to look at the actual unit performance to see if the stats make sense.

Take the Botroas versus Hastati example, with an armour rating of 1 the Hastati would be very vulnerable to missiles and individual unit combat. If they were not used en-mass to charge an opponent they would be slaughtered. By contrast this is very much the behaviour shown by the Botroas and actually supported by the offensive statistics. So while the armour value perhaps makes little sense in the context of the armour worn by the Hastati, it may well be the case that it exists purely to produce the desired combat behaviour and performance the unit is supposed to historically show. In this respect the game only has three defensive values to produce the directional defensive capability of a unit. The early scutum for example would be a far greater tool of defence than simply +4 to defence for front and left of a single man, but you would expect instead for the shield to be utilised more like a wall presented on all sides of the unit protecting the troops behind and being for all the world like row upon row of wooden barricades. I would therefore in this example consider the 7 armour rating to represent the defensive effect of each man in a maniple wielding a scutum. A large number of scutums in a maniple would inevitably boost the "armour" rating of the entire unit because of the sheer quantity and size of the shields being carried.

That might not be the reasoning, but the important factor is that the armour value of the Hastati allows them greater defence from multiple angles of attack, and therefore allows them to perform a line holding role.

I think when you get down to unit statistics, you cannot afford to try and equate armour worn to armour value or shields wielded to shield value, but instead must look for the statistics that generate the required combat behaviour effects. If unarmoured Hastati require a relatively high armour value in EB to actually produce the documented and known combat behaviour, then I would certainly do so rather than take statistical realism too far and destroy combat behaviour realism.

The final point is that perfect representation in the statistics would require that the fundamental game mechanics of combat are themselves realistic to the point where realistic values produce realistic outcomes. As things stand, while there may be discrepancies in values, and unarmoured units having armour values, the actual final result on the battlefield is something I am very happy with and it "looks right" to me in combat.